Self-similarity based single frame super-resolution

A method of generating a super-resolution image from a single frame, comprising:

• receiving, at a processor, first image data having a first resolution as a query image layer;
• acquiring, by the processor, second image data as a copy image layer;
• receiving at a processor, second image data having an original resolution as a second copy layer;
• generating, by the processor, at least two additional images layer as retrieval layers, the retrieval layers having a resolution lower than the copy layers, wherein the lower resolution corresponds to a copy layer resolution divided by a scaling factor;
• generating high resolution image data for a high resolution layer having a resolution higher than the first resolution, the generating including:
• selecting a pixel position in the high resolution layer;
• determining if a match is found between a query patch in the query layer, the query patch corresponding to the pixel position in the high resolution layer;
• when a match is found, determining if the query patch is biased towards a strong edge;
• when the query patch is biased towards a strong edge, copying image data from the second copy layer;
• and
• repeating the receiving, acquiring, generating at least one additional image layer, and generating high resolution image data through as many iterations as needed to achieve a desired resolution.